MATLAB: Different Values if K-means Clustring on same data.

clustringk-meansStatistics and Machine Learning Toolbox

I have been using matlab function of K-means clustring for making clusters of data. I happen to apply it on same data. But got wildly different results every time. I know the reason for this. But I need sugestions for overcoming this issue. Should I use some modified version of K-means or Should look for some other clustering technique?

K-means command which i used is "kmeans(Feature_Matrix,20,'Replicates',5,'emptyaction','singleton');

Best Answer

Kmeans can get stuck in local minima. By which I mean it is sensitive to initial centroid positions. You can specify a higher number of replicates to increase you chances of getting a global solution.

If you are interested in exploring other clustering algorithms, find all the supported ones here:

http://www.mathworks.com/machine-learning/

Related Solutions

MATLAB: K-means Clustering Result Always Changes

%generate some initial cluster centers according to some deterministic algorithm
%in this case, I construct a space-diagonal equally spaced, but choose your
%own algorithm
minc = min(cobat, 1);
maxc = max(cobat, 1);
nsamp = size(cobat,1);
initialcenters = repmat(minc, nsamp, 1) + bsxfun(@times, (0:nsamp-1).', (maxc - minc) ./ (nsamp-1));
%Once you have constructed the initial centers, cluster using those centers
[cidx ctrs] = kmeans(cobat, k, 'dist', 'sqEuclidean', 'start', initialcenters);

MATLAB: K-means without iteration

Not sure I follow exactly, but you could use grpstats to compute the coordinatewise means of data for each distinct value of T. You could use pdist2 to compute the distance from each data point to each centroid. You could use min to figure out which centroid is closest to each point (note the second output of the min command).

Best Answer

Related Solutions

MATLAB: K-means Clustering Result Always Changes

MATLAB: K-means without iteration

Related Question